1. Field of the Invention
The present invention relates to an information recording medium, an information recording method, an information recording apparatus and an information reproducing apparatus.
2. Description of the Related Art
An optical disk is a type of information recording medium which has a sector structure. In recent years, as the recording density and the capacity of an optical disk have been increased, it has become more important to ensure the reliability thereof. In order to ensure the reliability, an optical disk apparatus performs defect management in which a sector on the disk which cannot be used for recording/reproduction (hereinafter, referred to as a "defective sector") is replaced by another sector having a good condition. One standard for such defect management is ISO/IEC 10090 for 90 mm optical disks (hereinafter, referred to as the "ISO standard"), which is published from International Standards Organization (ISO).
As the first prior art example, an ECC block which is used by a DVD standard and the defect management method according to the ISO standard will be briefly described below.
FIG. 17 illustrates a physical structure of a disk 1. The disk 1 has a plurality of tracks 2 provided in the form of concentric circles or a spiral. Each of the tracks 2 is divided into a plurality of sectors 3. The disk 1 includes one or more disk information areas 4 and a data recording area 5.
The disk information area 4 stores various parameters needed to access the disk 1. In the example illustrated in FIG. 17, two disk information areas 4 are provided respectively along the inner and outer peripheries of the disk 1. The disk information area 4 along the inner periphery is also called a "lead-in" area, while the disk information area 4 along the outer periphery is also called a "lead-out" area.
Data is recorded/reproduced on/from the data recording area 5. Each sector 3 in the data recording area 5 is assigned an absolute address which is called a "physical sector number".
FIG. 18A illustrates a structure of an ECC (error correcting code) block which is a unit of error correcting code calculation. An ECC block contains main data (172 bytes.times.48 rows), an inner code parity PI obtained by calculating error correcting codes for each row (in the horizontal direction), and an outer code parity PO obtained by calculating error correcting codes for each column (in the vertical direction).
An error correction method using such inner and outer parities is generally called a "product code-based error correction method". The product code-based error correction method is an error correction method which is effective for both random errors and burst errors (a group of localized errors). For example, consider a case where some random errors occurred, as well as two rows of burst errors due to a scratch made on the disk 1. Most of such burst errors are correctable using the outer codes, because they are 2-byte errors in the vertical direction. A column with many random errors may not completely be corrected by outer codes. Some errors may remain after an error correction operation using outer codes. However, such remaining errors are in most cases correctable using inner codes. Even if some errors still remain after the error correction operation using inner codes, such errors can further be reduced by performing an error correction operation using outer codes again. By employing such product codes, DVDs realize a sufficient error correction capability while saving the parity redundancy. In other words, the capacity for user data is increased by such saving of the parity redundancy.
In a larger capacity DVD, each ECC block includes 16 sectors so as to realize both an increased error correction capability and a reduced redundancy. The ECC block illustrated in FIG. 18A includes only 4 sectors for the sake of simplicity.
FIG. 18B illustrates an arrangement of sectors included in an ECC block. The outer code parities PO of the ECC block are divided into rows and proportionally distributed among the sectors. As a result, each recording sector includes data of 182 bytes.times.13 rows.
An upper level control unit (this generally corresponds to a host computer) instructs an optical disk apparatus to record or reproduce data by sectors. When instructed to reproduce data from a sector, the optical disk apparatus reproduces an ECC block including the sector from the disk, performs error correction on the reproduced data, and returns only a portion of the data which corresponds to the designated sector. When instructed to record data on a sector, the optical disk apparatus reproduces an ECC block including the sector from the disk, performs error correction on the reproduced data, and replaces a portion of the data which corresponds to the designated sector with recording data which has been received from the upper level control unit. Then, the optical disk apparatus recalculates error correcting codes for the ECC block and adds them to the ECC block, before the ECC block including the designated sector is recorded on the disk. Particularly, such a recording operation is called a "read modified write" operation.
In the following description, a "block" means an ECC block as described above.
FIG. 19 illustrates an exemplary physical space of the disk 1 for use with the defect management method according to the ISO standard. The data recording area 5 includes a volume space 6 and a spare area 9.
The volume space 6 is managed by consecutive addresses, called "logical sector numbers". The volume space 6 includes a logical volume space 6a and logical volume structures 6b for storing information on the structure of the logical volume space 6a.
The spare area 9 includes at least one sector (for example, #1 spare block) which may be used in place of a defective sector if such a defective sector occurs in the volume space 6.
In the example illustrated in FIG. 19, a file A (indicated as "File-A" in FIG. 19) exists directly under a root directory (indicated as "ROOT" in FIG. 19). Among data blocks a to a included in the data extent of the root directory, the data block a is defective. The defective block a is replaced by #1 spare block in the spare area 9. Among data blocks d to g included in the data extent of the file A, the data block f is defective. The defective block f is replaced by #2 spare block in the spare area 9.
The replacement of each defective block by a spare block in the spare area 9 is registered in a secondary defect list ("SDL"). The SDL is stored in a defect management information area as a part of defect management information.
More recently, there is an attempt in the art to use a rewritable optical disk in a less expensive form of a bare disk with no cartridge, as a read-only optical disk. In view of the defect management, however, a bare disk is more likely to get fingerprints thereon, and the number of defective sectors may increase unexpectedly. Therefore, it is proposed in the art to use a dynamically expandable spare area rather than a fixed spare area.
Moreover, the increased capacity of an optical disk, along with the motion picture compression technique having been put into practical use, has paved the way to recording/reproduction of motion pictures on/from an optical disk. However, the conventional defect management method may not be suitable for such a motion picture application, in which real time processing is required. In particular, if a defective sector is replaced by a spare sector which is physically distant from the defective sector, it may take too much time to move the optical head to such a distant spare sector for ensuring the real time processing. Therefore, it has been proposed in the art to employ a new defect management method instead of the conventional method where a defective sector is replaced by a physically distant spare sector.
As the second prior art example, a proposed method for recording/reproducing AV data (i.e., audio video data) will be described below.
Each of FIGS. 20A and 20B illustrates an arrangement of AV data on a disk, which is suitable for AV data recording/reproduction. In FIGS. 20A and 20B, a suffix "h" denotes a hexadecimal number.
FIG. 20A illustrates an AV data arrangement where there is no defective sector. If there is no defective sector, the AV data including #1 data to #4 data can be recorded in sectors having consecutive logical sector numbers (LSN). Similarly, the AV data can be reproduced by reproducing the sectors having the consecutive logical sector numbers.
FIG. 20B illustrates an AV data arrangement where 16 sectors having logical sector numbers of n to n+0Fh are detected as defective sectors while recording data therein. In this case, the ECC block including the detected defective sector is skipped. As a result, #3 data is recorded in sectors having logical sector numbers of n+10h to n+1Fh, and #4 data is recorded in the following sectors having logical sector numbers of n+20h to n+2Fh. Such an operation of skipping sectors by ECC blocks is referred to as a "block skip".
FIG. 21 illustrates an exemplary physical space of the disk 1 which is suitable for AV data recording/reproduction.
In the example illustrated in FIG. 21, a file A (indicated as "File-A" in FIG. 21) containing AV data exists directly under a root directory (indicated as "ROOT" in FIG. 21). Among data blocks a to c included in the data extent of the root directory, the data block a is defective. The defective block a is replaced by #1 spare block in the spare area 9. It is assumed that a defective block f is detected while recording the AV data extent of the file A in an area provided for the AV data extent. In such a case, the defective block f is skipped. As a result, the AV data extent of the file A is recorded while being divided into an AV data extent I (including the data blocks d and e) and another AV data extent II (including the data blocks g and h).
The replacement of the defective block c by #1 spare block in the spare area 9 is registered in the SDL. However, the defective block f is not registered in the SDL because the defective block f was only skipped, and the defective block f is not replaced by a spare block (no spare block has even been allocated thereto).
However, there is a problem associated with the presence of such a defective block which is not registered in the SDL. The problem will be described below with reference to FIGS. 22A to 22C.
FIG. 22A illustrates a normally recorded ECC block. The ECC block is recorded over a plurality of sectors. Each sector begins with an ID containing the physical sector number of the sector, etc. Data is recorded in the area following the ID. The data is obtained by adding error correcting codes to main data and further interleaving the main data having the error correcting codes added thereto (see FIG. 18).
FIG. 22B illustrates an ECC block for which an overwrite operation failed. When the ECC block illustrated in FIG. 22A is overwritten with new data, new error correcting codes are calculated according to the new main data, and added to the ECC block. In the example illustrated in FIG. 22B, however, the third sector has a defective ID. Therefore, the first two sectors are overwritten with data of a new ECC block, while the other two sectors remain to have the data of an old ECC block.
FIG. 22C illustrates the structure of reproduced data from the ECC block for which the overwrite operation failed. When the four sectors illustrated in FIG. 22B are reproduced, the new data and the old data are mixed with each other (in FIG. 22C, the new data and the old data are hatched in different directions). This means that an error correction always fails in the vertical direction using the outer code parity P0.
As can be appreciated from the description above, a block for which a recording operation even once failed becomes a block from which data cannot be reproduced. The read modified write operation is required to record data in some sectors of this block. However, a read modified write operation for such an unreproducible block will always fail. Thus, this block becomes a block on which data can no longer be recorded. Such a block cannot later be replaced by a replacement block because data to be transferred to the replacement block cannot be reproduced from the block, as in the read modified write operation.
If the dynamically expandable spare area was used with the ISO standard defect management method which is designed for use with a fixed-size spare area, the spare area may temporarily be exhausted (i.e., no available spare area), which would never happen in the conventional techniques. There is no method proposed in the art to manage a defective block which is detected while the spare area is temporarily exhausted. Since a read modified write operation for such an unmanaged defective block fails, data cannot be recorded by sectors in the defective block. Also when recording/reproducing AV data on/from the disk, a read modified write operation for a skipped defective block fails, thereby experiencing the same problem as just described above.